In this paper, we introduce DenseControl, a novel pipeline for generating dense crowd images. Specifically, DenseControl meticulously positions and sizes each generated instance to align precisely with the predefined coordinates and scales. Based on this, we further allow for control over the background, style, and attributes of instances. The motivation behind DenseControl stems from the observation of two main challenges in synthesizing crowd images: controlling signal embedding and maintaining topological integrity when imparting instance scale guidance. To address these, we first introduce the Isolated Object Embedding (IOE) map, a novel representation that facilitates spatial location control while mitigating the difficulties associated with learning projections for model. Secondly, we propose an Implicit Scale Embedding (ISE) strategy that seamlessly integrates with the IOE map to encode precise scale information. To further enhance the efficacy of combining ISE with the IOE map, we incorporate a Position Shortcut mechanism that enhances cross-attention to alleviate projection challenges. We evaluate DenseControl through two lenses: synthesis quality and applicability in latent applications. Experiments across different control conditions demonstrate DenseControl achieves state-of-the-art results in dense crowd image synthesis. Furthermore, we showcase applications in augmenting crowd analysis under data scarcity, transfer learning, and weather generalization scenes, to highlight the practical utility of DenseControl. The codebase will be released.

Crowd localization plays a crucial role in visual scene understanding towards predicting each pedestrian location in a crowd, thus being applicable to various downstream tasks. However, existing approaches suffer from significant performance degradation due to discrepancies in head scale distributions (scale shift) between training and testing data, a challenge known as domain generalization (DG). This paper aims to comprehend the nature of scale shift within the context of domain generalization for crowd localization models. To this end, we address four critical questions: (i) How does scale shift influence crowd localization in a DG scenario? (ii) How can we quantify this influence? (iii) What causes this influence? (iv) How to mitigate the influence? Initially, we conduct a systematic examination of how crowd localization performance varies with different levels of scale shift. Then, we establish a benchmark, ScaleBench, and reproduce 20 advanced DG algorithms to quantify the influence. Through extensive experiments, we demonstrate the limitations of existing algorithms and underscore the importance and complexity of scale shift, a topic that remains insufficiently explored. To deepen our understanding, we provide a rigorous theoretical analysis on scale shift. Building on these insights, we further propose an effective algorithm called Causal Feature Decomposition and Anisotropic Processing (Catto) to mitigate the influence of scale shift in DG settings. Later, we also provide extensive analytical experiments, revealing four significant insights for future research. Our results emphasize the importance of this novel and applicable research direction, which we term Scale Shift Domain Generalization. The presented novel dataset, algorithms will be released at https://github.com/wjc2830/ScaleBench.git .

While language models (LMs) paired with residual vector quantization (RVQ) tokenizers have shown promise in text-to-audio (T2A) generation, they still lag behind diffusion-based models by a non-trivial margin. We identify a critical dilemma underpinning this gap: incorporating more RVQ layers improves audio reconstruction fidelity but exceeds the generation capacity of conventional LMs. To address this, we first analyze RVQ dynamics and uncover two key limitations: 1) orthogonality of features across RVQ layers hinders effective LMs training, and 2) descending semantic richness in tokens from deeper RVQ layers exacerbates exposure bias during autoregressive decoding. Based on these insights, we propose Siren, a novel LM-based framework that employs multiple isolated transformers with causal conditioning and anti-causal alignment via reinforcement learning. Extensive experiments demonstrate that Siren outperforms both existing LM-based and diffusion-based T2A systems, achieving state-of-the-art results. By bridging the representational strengths of LMs with the fidelity demands of audio synthesis, our approach repositions LMs as competitive contenders against diffusion models in T2A tasks. Moreover, by aligning audio representations with linguistic structures, Siren opens a promising pathway toward unified multi-modal generation frameworks.

Video-to-audio generation is essential for synthesizing realistic audio tracks that synchronize effectively with silent videos. Following the perspective of extracting essential signals from videos that can precisely control the mature text-to-audio generative diffusion models, this paper presents how to balance the representation of mel-spectrograms in terms of completeness and complexity through a new approach called Mel Quantization-Continuum Decomposition (Mel-QCD). We decompose the mel-spectrogram into three distinct types of signals, employing quantization or continuity to them, we can effectively predict them from video by a devised video-to-all (V2X) predictor. Then, the predicted signals are recomposed and fed into a ControlNet, along with a textual inversion design, to control the audio generation process. Our proposed Mel-QCD method demonstrates state-of-the-art performance across eight metrics, evaluating dimensions such as quality, synchronization, and semantic consistency.

The Segment Anything Model (SAM) and its family of models have made significant strides in open-set, prompt-driven instance segmentation. However, some closed-source SAM family models often face ethical, copyright, or commercial restrictions, limiting their accessibility and further personalized adaptation. To overcome these limitations, we introduce MapSAM, a model-agnostic, personalized plugin for SAM family models. MapSAM features a lightweight threshold learner that enables nuanced post-hoc processing of confidence maps, leading to improved segmentation accuracy. By leveraging mask-focused learning, our approach determines pixel-wise and hardness-aware thresholds, allowing for more effective adaptation to diverse datasets. Furthermore, we critically examine the limitations of the commonly used Dice loss, which can overlook sample hardness when allocating penalties. We theoretically demonstrate that the Mean Squared Error (MSE) loss complements Dice loss by providing a stronger focus on sample hardness. Through extensive experiments on seven diverse datasets using multiple SAM family models, we validate the effectiveness of MapSAM in achieving superior segmentation results, particularly in challenging domains. Our findings open up new avenues for personalized, open-set instance segmentation across various application areas, leveraging any closed-source SAM family model. Code will be available at https://github.com/wjc2830/MapSAM.git.

Data pruning aims to obtain lossless performances with less overall cost. A common approach is to filter out samples that make less contribution to the training. This could lead to gradient expectation bias compared to the original data. To solve this problem, we propose \textbf{InfoBatch}, a novel framework aiming to achieve lossless training acceleration by unbiased dynamic data pruning. Specifically, InfoBatch randomly prunes a portion of less informative samples based on the loss distribution and rescales the gradients of the remaining samples to approximate the original gradient. As a plug-and-play and architecture-agnostic framework, InfoBatch consistently obtains lossless training results on classification, semantic segmentation, vision pertaining, and instruction fine-tuning tasks. On CIFAR10/100, ImageNet-1K, and ADE20K, InfoBatch losslessly saves 40\% overall cost. For pertaining MAE and diffusion model, InfoBatch can respectively save 24.8\% and 27\% cost. For LLaMA instruction fine-tuning, InfoBatch is also able to save 20\% cost and is compatible with coreset selection methods. The code is publicly available at \href{https://github.com/henryqin1997/InfoBatch}{github.com/NUS-HPC-AI-Lab/InfoBatch}.

In the field of human-centric personalized image generation, the adapter-based method obtains the ability to customize and generate portraits by text-to-image training on facial data. This allows for identity-preserved personalization without additional fine-tuning in inference. Although there are improvements in efficiency and fidelity, there is often a significant performance decrease in test following ability, controllability, and diversity of generated faces compared to the base model. In this paper, we analyze that the performance degradation is attributed to the failure to decouple identity features from other attributes during extraction, as well as the failure to decouple the portrait generation training from the overall generation task. To address these issues, we propose the Face Adapter with deCoupled Training (FACT) framework, focusing on both model architecture and training strategy. To decouple identity features from others, we leverage a transformer-based face-export encoder and harness fine-grained identity features. To decouple the portrait generation training, we propose Face Adapting Increment Regularization~(FAIR), which effectively constrains the effect of face adapters on the facial region, preserving the generative ability of the base model. Additionally, we incorporate a face condition drop and shuffle mechanism, combined with curriculum learning, to enhance facial controllability and diversity. As a result, FACT solely learns identity preservation from training data, thereby minimizing the impact on the original text-to-image capabilities of the base model. Extensive experiments show that FACT has both controllability and fidelity in both text-to-image generation and inpainting solutions for portrait generation.

Dimension reduction (DR) is commonly utilized to capture the intrinsic structure and transform high-dimensional data into low-dimensional space while retaining meaningful properties of the original data. It is used in various applications, such as image recognition, single-cell sequencing analysis, and biomarker discovery. However, contemporary parametric-free and parametric DR techniques suffer from several significant shortcomings, such as the inability to preserve global and local features and the poor generalisation performance. On the other hand, regarding explainability, it is crucial to comprehend the embedding process, especially the contribution of each part to the embedding process, while understanding how each feature affects the embedding results that identify critical components and help diagnose the embedding process. To address these problems, we have developed a deep neural network method called DMT-EV, which provides not only excellent performance in structural maintainability but also explainability to the DR therein. DMT-EV starts with data augmentation and a manifold-based loss function to improve embedding performance. The explanation is based on saliency maps and aims to examine the trained DMT-EV parameters and contributions of components during the embedding process. The proposed techniques are integrated with a visual interface to help the user to adjust DMT-EV to achieve better DR performance and explainability. The interactive visual interface makes it easier to illustrate the data features, compare different DR techniques, and investigate DR. An in-depth experimental comparison shows that DMT-EV consistently outperforms the state-of-the-art methods in both performance measures and explainability.

Recently, subject-driven generation has garnered significant interest due to its ability to personalize text-to-image generation. Typical works focus on learning the new subject's private attributes. However, an important fact has not been taken seriously that a subject is not an isolated new concept but should be a specialization of a certain category in the pre-trained model. This results in the subject failing to comprehensively inherit the attributes in its category, causing poor attribute-related generations. In this paper, motivated by object-oriented programming, we model the subject as a derived class whose base class is its semantic category. This modeling enables the subject to inherit public attributes from its category while learning its private attributes from the user-provided example. Specifically, we propose a plug- and-play method, Subject-Derived regularization (SuDe). It constructs the base-derived class modeling by constraining the subject- driven generated images to semantically belong to the subject's category. Extensive experiments under three baselines and two backbones on various subjects show that our SuDe enables imaginative attribute-related generations while maintaining subject fidelity. For the codes, please refer to FaceChain.

Annotation-scarce semantic segmentation aims to obtain meaningful pixel-level discrimination with scarce or even no manual annotations, of which the crux is how to utilize unlabeled data by pseudo-label learning. Typical works focus on ameliorating the error-prone pseudo-labeling, e.g., only utilizing high-confidence pseudo labels and filtering low-confidence ones out. But we think differently and resort to exhausting informative semantics from multiple probably correct candidate labels. This brings our method the ability to learn more accurately even though pseudo labels are unreliable. In this paper, we propose Adaptive Fuzzy Positive Learning (A-FPL) for correctly learning unlabeled data in a plug-and-play fashion, targeting adaptively encouraging fuzzy positive predictions and suppressing highly probable negatives. Specifically, A-FPL comprises two main components: (1) Fuzzy positive assignment (FPA) that adaptively assigns fuzzy positive labels to each pixel, while ensuring their quality through a T-value adaption algorithm (2) Fuzzy positive regularization (FPR) that restricts the predictions of fuzzy positive categories to be larger than those of negative categories. Being conceptually simple yet practically effective, A-FPL remarkably alleviates interference from wrong pseudo labels, progressively refining semantic discrimination. Theoretical analysis and extensive experiments on various training settings with consistent performance gain justify the superiority of our approach. Codes are at A-FPL.

Deep learning benefits from the growing abundance of available data. Meanwhile, efficiently dealing with the growing data scale has become a challenge. Data publicly available are from different sources with various qualities, and it is impractical to do manual cleaning against noise and redundancy given today's data scale. There are existing techniques for cleaning/selecting the collected data. However, these methods are mainly proposed for offline settings that target one of the cleanness and redundancy problems. In practice, data are growing exponentially with both problems. This leads to repeated data curation with sub-optimal efficiency. To tackle this challenge, we propose InfoGrowth, an efficient online algorithm for data cleaning and selection, resulting in a growing dataset that keeps up to date with awareness of cleanliness and diversity. InfoGrowth can improve data quality/efficiency on both single-modal and multi-modal tasks, with an efficient and scalable design. Its framework makes it practical for real-world data engines.

Common space learning, associating semantic and visual domains in a common latent space, is essential to transfer knowledge from seen classes to unseen ones on Zero-Shot Learning (ZSL) realm. Existing methods for common space learning rely heavily on structure alignment due to the heterogeneous nature between semantic and visual domains, but the existing design is sub-optimal. In this paper, we utilize persistent homology to investigate geometry structure alignment, and observe two following issues: (i) The sampled mini-batch data points present a distinct structure gap compared to global data points, thus the learned structure alignment space inevitably neglects abundant and accurate global structure information. (ii) The latent visual and semantic space fail to preserve multiple dimensional geometry structure, especially high dimensional structure information. To address the first issue, we propose a Topology-guided Sampling Strategy (TGSS) to mitigate the gap between sampled and global data points. Both theoretical analyses and empirical results guarantee the effectiveness of the TGSS. To solve the second issue, we introduce a Topology Alignment Module (TAM) to preserve multi-dimensional geometry structure in latent visual and semantic space, respectively. The proposed method is dubbed TopoZero. Empirically, our TopoZero achieves superior performance on three authoritative ZSL benchmark datasets.

Face detection (FD) has achieved remarkable success over the past few years, yet, these leaps often arrive when consuming enormous computation costs. Moreover, when considering a realistic situation, i.e., building a lightweight face detector under a computation-scarce scenario, such heavy computation cost limits the application of the face detector. To remedy this, several pioneering works design tiny face detectors through off-the-shelf neural architecture search (NAS) technologies, which are usually applied to the classification task. Thus, the searched architectures are sub-optimal for the face detection task since some design criteria between detection and classification task are different. As a representative, the face detection backbone design needs to guarantee the stage-level detection ability while it is not required for the classification backbone. Furthermore, the detection backbone consumes a vast body of inference budgets in the whole detection framework. Considering the intrinsic design requirement and the virtual importance role of the face detection backbone, we thus ask a critical question: How to employ NAS to search FD-friendly backbone architecture? To cope with this question, we propose a distribution-dependent stage-aware ranking score (DDSAR-Score) to explicitly characterize the stage-level expressivity and identify the individual importance of each stage, thus satisfying the aforementioned design criterion of the FD backbone. Based on our proposed DDSAR-Score, we conduct comprehensive experiments on the challenging Wider Face benchmark dataset and achieve dominant performance across a wide range of compute regimes. In particular, compared to the tiniest face detector SCRFD-0.5GF, our method is +2.5 % better in Average Precision (AP) score when using the same amount of FLOPs. The code is avaliable at https://github.com/ly19965/EasyFace/tree/ master/face_project/face_detection/DamoFD.

Data uncertainty is commonly observed in the images for face recognition (FR). However, deep learning algorithms often make predictions with high confidence even for uncertain or irrelevant inputs. Intuitively, FR algorithms can benefit from both the estimation of uncertainty and the detection of out-of-distribution (OOD) samples. Taking a probabilistic view of the current classification model, the temperature scalar is exactly the scale of uncertainty noise implicitly added in the softmax function. Meanwhile, the uncertainty of images in a dataset should follow a prior distribution. Based on the observation, a unified framework for uncertainty modeling and FR, Random Temperature Scaling (RTS), is proposed to learn a reliable FR algorithm. The benefits of RTS are two-fold. (1) In the training phase, it can adjust the learning strength of clean and noisy samples for stability and accuracy. (2) In the test phase, it can provide a score of confidence to detect uncertain, low-quality and even OOD samples, without training on extra labels. Extensive experiments on FR benchmarks demonstrate that the magnitude of variance in RTS, which serves as an OOD detection metric, is closely related to the uncertainty of the input image. RTS can achieve top performance on both the FR and OOD detection tasks. Moreover, the model trained with RTS can perform robustly on datasets with noise. The proposed module is light-weight and only adds negligible computation cost to the model.

Recent advancement in personalized image generation have unveiled the intriguing capability of pre-trained text-to-image models on learning identity information from a collection of portrait images. However, existing solutions are vulnerable in producing truthful details, and usually suffer from several defects such as (i) The generated face exhibit its own unique characteristics, facial shape and facial feature positioning may not resemble key characteristics of the input, and (ii) The synthesized face may contain warped, blurred or corrupted regions. In this paper, we present FaceChain, a personalized portrait generation framework that combines a series of customized image-generation model and a rich set of face-related perceptual understanding models (, face detection, deep face embedding extraction, and facial attribute recognition), to tackle aforementioned challenges and to generate truthful personalized portraits, with only a handful of portrait images as input. Concretely, we inject several SOTA face models into the generation procedure, achieving a more efficient label-tagging, data-processing, and model post-processing compared to previous solutions, such as DreamBooth , InstantBooth , or other LoRA-only approaches . Besides, based on FaceChain, we further develop several applications to build a broader playground for better showing its value, including virtual try on and 2D talking head. We hope it can grow to serve the burgeoning needs from the communities. Note that this is an ongoing work that will be consistently refined and improved upon. FaceChain is open-sourced under Apache-2.0 license at .

Unsupervised domain adaptation (UDA) has proven to be highly effective in transferring knowledge from a label-rich source domain to a label-scarce target domain. However, the presence of additional novel categories in the target domain has led to the development of open-set domain adaptation (ODA) and universal domain adaptation (UNDA). Existing ODA and UNDA methods treat all novel categories as a single, unified unknown class and attempt to detect it during training. However, we found that domain variance can lead to more significant view-noise in unsupervised data augmentation, which affects the effectiveness of contrastive learning (CL) and causes the model to be overconfident in novel category discovery. To address these issues, a framework named Soft-contrastive All-in-one Network (SAN) is proposed for ODA and UNDA tasks. SAN includes a novel data-augmentation-based soft contrastive learning (SCL) loss to fine-tune the backbone for feature transfer and a more human-intuitive classifier to improve new class discovery capability. The SCL loss weakens the adverse effects of the data augmentation view-noise problem which is amplified in domain transfer tasks. The All-in-One (AIO) classifier overcomes the overconfidence problem of current mainstream closed-set and open-set classifiers. Visualization and ablation experiments demonstrate the effectiveness of the proposed innovations. Furthermore, extensive experiment results on ODA and UNDA show that SAN outperforms existing state-of-the-art methods.

: Recent advancement in personalized image generation have unveiled the intriguing capability of pre-trained text-to-image models on learning identity information from a collection of portrait images. However, existing solutions can be vulnerable in producing truthful details, and usually suffer from several defects such as (i) The generated face exhibit its own unique characteristics, i.e . facial shape and facial feature positioning may not re-semble key characteristics of the input, and (ii) The synthesized face may contain warped, blurred or corrupted regions. In this paper, we present FaceChain, a personalized portrait generation framework that combines a series of customized image-generation model and a rich set of face-related perceptual understanding models ( e.g ., face detection, deep face embedding extraction, and facial attribute recognition), to tackle aforementioned challenges and to generate truthful personalized portraits, with only a handful of portrait images as input. Concretely, we inject several SOTA face models into the generation procedure, achieving a more efficient label-tagging, data-processing, and model post-processing compared to previous solutions, such as DreamBooth [9] , InstantBooth [11] , or other LoRA-only approaches [3] . Through the development of FaceChain, we have identified several potential directions to accelerate development of Face/Human-Centric AIGC research and application. We have designed FaceChain as a framework comprised of pluggable components that can be easily adjusted to accommodate different styles and personalized needs. We hope it can grow to serve the burgeoning needs from the communities. Note that this is an ongoing work that will be consistently refined and improved upon. FaceChain is open-sourced under Apache-2.0 license at https://github.com/modelscope/facechain .

Face detection (FD) has achieved remarkable success over the past few years, yet, these leaps often arrive when consuming enormous computation costs. Moreover, when considering a realistic situation, i.e., building a lightweight face detector under a computation-scarce scenario, such heavy computation cost limits the application of the face detector. To remedy this, several pioneering works design tiny face detectors through off-the-shelf neural architecture search (NAS) technologies, which are usually applied to the classification task. Thus, the searched architectures are sub-optimal for the face detection task since some design criteria between detection and classification task are different. As a representative, the face detection backbone design needs to guarantee the stage-level detection ability while it is not required for the classification backbone. Furthermore, the detection backbone consumes a vast body of inference budgets in the whole detection framework. Considering the intrinsic design requirement and the virtual importance role of the face detection backbone, we thus ask a critical question: How to employ NAS to search FD-friendly backbone architecture? To cope with this question, we propose a distribution-dependent stage-aware ranking score (DDSAR-Score) to explicitly characterize the stage-level expressivity and identify the individual importance of each stage, thus satisfying the aforementioned design criterion of the FD backbone. Based on our proposed DDSAR-Score, we conduct comprehensive experiments on the challenging Wider Face benchmark dataset and achieve dominant performance across a wide range of compute regimes. In particular, compared to the tiniest face detector SCRFD-0.5GF, our method is +2.5 % better in Average Precision (AP) score when using the same amount of FLOPs. The code is avaliable at https://github.com/ly19965/FaceMaas/tree/master/face_project/face_detection/DamoFD.

Domain adaptation (DA) transfers knowledge from label-rich domains to new domains where labels are scarce to address the problem of generalization of deep neural networks in new domains. Universal domain adaptation (UNDA) assumes the label distributions of labeled and unlabeled data are different and unknowable. In this paper, we concentrate on solving the noise problem on the UNDA problem based on contrastive learning (CL), which includes view noise in data augmentation and label noise in the classifier training. The domain differences in UNDA amplify the noise in the view of data augmentation, resulting in data augmentation schemes that apply to all domains being challenging to find. In addition, the mainstream UNDA classifiers combine closed-set classifiers with open-set classifiers; insufficient competition among open-set classifiers leads to overconfidence, which results in incredible sensitivity to noise in labeled data. Therefore, we propose Noise-Resistant Soft Contrastive Learning (NSCL) addresses the above issues. Firstly, we propose a soft contrast learning loss to avoid the over-response of typical CL loss to noisy samples, thus enabling data augmentation to improve the performance of UNDA further. Secondly, we design an all-in-one (AIO) classifier to improve the robustness of noisy labels while introducing multi-category unknown class competition. Extensive experimental results on UNDA and openset DA demonstrate the advantages of NSCL over existing methods, especially in downstream tasks such as classification and visualization.

Unsupervised domain adaptation (UDA) aims to transfer the knowledge from a labeled source domain to an unlabeled target domain. Typically, to guarantee desirable knowledge transfer, aligning the distribution between source and target domain from a global perspective is widely adopted in UDA. Recent researchers further point out the importance of local-level alignment and borrow the experience from Optimal Transport (OT) theory to construct instance-pair alignment. However, existing OT-based algorithms are limited to resolve class imbalance challenge and require a huge computation cost when considering a large-scale training situation. In this paper, we address these two issues by proposing a Clustering-based Optimal Transport (COT) algorithm, which formulates the alignment procedure as an Optimal Transport problem by capturing the fine-grained attribute alignment. Concretely, COT innovatively designs the loss derived from discrete Kantorovich dual form to construct a mapping between clustering centers in source and target domain, which simultaneously eliminates the negative effect brought by class imbalance and reduces the computation cost on the basis of theoretical analysis. Finally, our COT together with some previous UDA methods achieve superior performance on several benchmarks.